The wigner distribution in speech processing applications